Model Parameters Research Articles

Phenology- and light-based parameterization of stomatal conductance model for urban woody species in northern China

Surface ozone (O3) poses a significant threat to urban vegetation health, and assessing the O3 risk across woody species is of vital importance for maintaining the health of urban infrastructure. In the present study, Jarvis-type stomatal conductance model was parameterized for ten urban species in northern China. Incorporating the effects of time of day and diurnal O3 concentration significantly enhanced the model performance. For different plant functional types (greening trees, greening shrubs, and orchard-grown trees), three parameterizations were established to estimate stomatal O3 uptake (POD1, phytotoxic O3 dose over an hourly threshold of 1 nmol m-2 s-1). The differences in POD1 between greening trees and shrubs were primarily due to the difference in their stomatal sensitivity to light. Orchard-grown trees displayed the lowest O3 removal capacity (lowest value of POD1) because of their shorter growing season despite of high stomatal conductance. These results indicated that plant phenology and light responsiveness determined stomatal O3 uptake, and the three parameterizations developed here could be applicable to various urban species in northern regions. Among climatic factors for O3 risk assessment, O3 concentration was the most important factor determining annual variation of POD1, which was primarily driven by air temperature. However, when O3 pollution decreased, O3 concentration exhibited less dependence on temperature and more dependence on light. These findings provide crucial insights for urban policy-makers and environmental scientists aiming to mitigate O3 pollution effects and enhance urban vegetation health.

Allometric Comparison of Viral Dynamics in the Nasal Cavity-Nasopharyngeal Mucus Layer of Human and Rhesus Monkey by CFD-HCD Approach

Background and ObjectiveViral respiratory infections stand as a considerable global health concern, presenting significant risks to the health of both humans and animals. This study aims to conduct a preliminary analysis of the time series of viral load in the nasal cavity-nasopharynx (NC-NP) of the human and rhesus macaque (RM). MethodsTaking into account the random uniform distribution of virus-laden droplets with a diameter of 10μm in the mucus layer, this study applies the computational fluid dynamics-host cell dynamics (CFD-HCD) method to 3D-shell NC-NP models of human and RM, analyzing the impact of initial distribution of droplets on the viral dynamics of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), estimating parameters in the HCD model based on experimental data, integrating them into simulations to predict the time series of viral load and cell counts, and being visualized. The reproductive number (R0) are calculated to determine the occurrence of infection. The study also considers cross-parameter combinations and cross-experimental datasets to explore potential correlations between the human and RM. ResultsThe research findings indicate that the uniform distribution of virus-laden droplets throughout the whole NC-NP models of human and RM is reasonable for simulating and predicting viral dynamics. The visualization results offer dynamic insights into virus infection over a period of 20 days. Studies involving parameter and dataset exchanges between the two species underscore certain similarities in predicting virus infections between the human and RM. ConclusionsThis study lays the groundwork for further exploration into the parallels and distinctions in respiratory virus dynamics between humans and RMs, thus aiding in making more informed decisions in research and experimentation.

Population pharmacokinetic modeling of sufentanil in adult Korean patients undergoing cardiopulmonary bypass surgery.

Sufentanil is frequently used as an anesthetic agent in cardiac surgery owing to its cardiovascular safety and favorable pharmacokinetics. However, the pharmacokinetics profiles of sufentanil in patients undergoing cardiopulmonary bypass (CPB) surgery remain less understood, which is crucial for achieving the desired level of anesthesia and mitigating surgical complications. Therefore, this study aimed to develop a population pharmacokinetic model of sufentanil in patients undergoing CPB surgery and elucidate the clinical factors affecting its pharmacokinetic profile. Adult patients who underwent cardiac surgery with CPB and were administered sufentanil for anesthesia were enrolled. Arterial blood samples were collected to quantify plasma concentrations of sufentanil and clinical laboratory parameters, including inflammatory cytokines. A population pharmacokinetic model was established using nonlinear mixed-effects modeling. Simulations were performed using the pharmacokinetic parameters of the final model. Overall, 20 patients were included in the final analysis. Sufentanil pharmacokinetics were modeled using a two-compartment model, accounting for CPB effects. Sufentanil clearance increased 2.80-fold during CPB and warming phases, while the central compartment volume increased 2.74-fold during CPB. CPB was a significant covariate affecting drug clearance and distribution volume. No other significant covariates were identified despite increased levels of the inflammatory cytokines, including IL-6, IL-8, and TNF-α during CPB. The simulation indicated a 30 μg loading dose and 40 μg/h maintenance infusion for target-controlled infusion. Additionally, a bolus dose of 60 μg was added at CPB initiation to adjust for exposure changes during this phase. Considering the target sufentanil concentrations, a uniform dosing regimen was acceptable for effective analgesia.

Spiking Tucker Fusion Transformer for Audio-Visual Zero-Shot Learning.

The spiking neural networks (SNNs) that efficiently encode temporal sequences have shown great potential in extracting audio-visual joint feature representations. However, coupling SNNs (binary spike sequences) with transformers (float-point sequences) to jointly explore the temporal-semantic information still facing challenges. In this paper, we introduce a novel Spiking Tucker Fusion Transformer (STFT) for audio-visual zero-shot learning (ZSL). The STFT leverage the temporal and semantic information from different time steps to generate robust representations. The time-step factor (TSF) is introduced to dynamically synthesis the subsequent inference information. To guide the formation of input membrane potentials and reduce the spike noise, we propose a global-local pooling (GLP) which combines the max and average pooling operations. Furthermore, the thresholds of the spiking neurons are dynamically adjusted based on semantic and temporal cues. Integrating the temporal and semantic information extracted by SNNs and Transformers are difficult due to the increased number of parameters in a straightforward bilinear model. To address this, we introduce a temporal-semantic Tucker fusion module, which achieves multi-scale fusion of SNN and Transformer outputs while maintaining full second-order interactions. Our experimental results demonstrate the effectiveness of the proposed approach in achieving state-of-the-art performance in three benchmark datasets. The harmonic mean (HM) improvement of VGGSound, UCF101 and ActivityNet are around 15.4%, 3.9%, and 14.9%, respectively.

Distinct sources of uncertainty in simulations of the ocean biological carbon pump at different depths

Model uncertainty in simulating the biological carbon pump was quantified and partitioned using 14 models from the Coupled Model Intercomparison Project Phase 6. Uncertainty increases with depth. On the global scale, uncertainty in carbon export dominates above 900 m and uncertainty in transfer efficiency below. Reducing model uncertainty in carbon export and transfer efficiency offers similar benefits for understanding century-scale carbon sequestration and climate. These models produce three different qualitative patterns in transfer efficiency: one where it is globally homogenous and two opposite latitudinal patterns due to different model structures and parameters. The exponent b of the Martin curve, which has long been used to compare different representations of transfer efficiency, is shown here to underestimate uncertainty in transfer efficiency. This highlights the importance of using vertical profiles of carbon flux rather than the single exponent b in model validation and intercomparison exercises.

Comprehensive Production Index Prediction Using Dual-Scale Deep Learning in Mineral Processing.

In mineral processing, the dynamic nature of industrial data poses challenges for decision-makers in accurately assessing current production statuses. To enhance the decision-making process, it is crucial to predict comprehensive production indices (CPIs), which are influenced by both human operators and industrial processes, and demonstrate a strong dual-scale property. To improve the accuracy of CPIs' prediction, we introduce the high-frequency (HF) unit and low-frequency (LF) unit within our proposed dual-scale deep learning (DL) network. This architecture enables the exploration of nonlinear dynamic mapping in dual-scale industrial data. By integrating the Cloud-Edge collaboration mechanism with DL, our training strategy mitigates the dominance of HF data and guides networks to prioritize different frequency information. Through self-tuning training via Cloud-Edge collaboration, the optimal model structure and parameters on the cloud server are adjusted, with the edge model self-updating accordingly. Validated through online industrial experiments, our method significantly enhances CPIs' prediction accuracy compared to the baseline approaches.

A personal thermal comfort model based on causal artificial intelligence: a physiological sensor-enabled causal identifiability.

Personal thermal comfort affects occupants' health, well-being, and productivity. Its satisfaction is subjective, based on individual characteristics and dynamic environments, and challenging to understand, requiring predicted outcomes and explanations of how and why the outcomes happen. This research fulfills this concern using a personal thermal comfort model based on causal artificial intelligence. It encodes personal thermal comfort satisfaction based on a new causal-and-effect framework to connect the human mind and environmental factors. Random variables encode relevant factors, and the structural causal model performs cause-and-effect relationships. The do-calculus (e.g., d-separated and d-connected) draws the common sense of the model based on causal structure representation, contributing to a human-intelligent understanding. A directed acyclic graph and exact-inference-based variable elimination quantify the model parameters based on real-world observational data. The strength of causal relationships is verified based on causal odd ratio, causal sensitivity, and causal impact. The results highlight that our proposed model can encode physiological and physical factors to predict and explain personal thermal comfort satisfaction. It can predict and explain such satisfaction reasonably and robustly, converging to human-like interpretation. It can be applied to intelligent systems to understand personal thermal comfort.

The role of pressure anisotropy on quark stars in gravity’s rainbow

This work is seeking for the existence of stable quark stars (QSs) in the framework of a modified theory of gravity known as gravity’s rainbow. This modification comes from the fact that the geometry of spacetime depends on the energy of the test particle. We solve numerically the modified TOV equations and present the mass–radius (M–R) diagram for quark matter equations of state. To constrain the allowed values of the model parameters, we use current astrophysical measurements of the masses and radii of neutron stars. Finally, we investigate the dynamical stability of the hydrostatic equilibrium equations in gravity’s rainbow by analyzing the static stability, adiabatic index, and sound velocity profiles.

Robust ordinal regression for subsets comparisons with interactions

This paper is devoted to a robust ordinal method for learning the preferences of a decision maker between subsets. The decision model, derived from Fishburn and LaValle (1996) and whose parameters we learn, is general enough to be compatible with any strict weak order on subsets, thanks to the consideration of possible interactions between elements. Moreover, we accept not to predict some preferences if the available preference data are not compatible with a reliable prediction. A predicted preference is considered reliable if all the simplest models (Occam’s razor) explaining the preference data agree on it. Following the robust ordinal regression methodology, our predictions are based on an uncertainty set encompassing the possible values of the model parameters. We define a new ordinal dominance relation between subsets and design a procedure to determine whether this dominance relation holds. Numerical tests are provided on synthetic and real-world data to evaluate the richness and reliability of the preference predictions made.

Capacity fading knee-point recognition method and life prediction for lithium-ion batteries using segmented capacity degradation model

Aiming at the problem that the cycle life of the battery is hard to be accurately estimated, a segmented capacity degradation model is established based on the trend of capacity degradation rate. By applying a genetic algorithm (GA) to optimize the model parameters, GA-Weibull and GA-SVR (support vector regression) degradation models and their corresponding segmented models are established, and the battery life is calculated. Furthermore, a method taking the RMSE as the target is put forward to calculate the knee-point of capacity degradation. Finally, the life is predicted by segmented models. The results show that compared with unsegmented models, describing the process of capacity degradation by segmented models reduces the error notably, makes the determination coefficient closer to 1, and improves the model accuracy effectively. Later, the proposed method makes the fitting error of the segmented model smaller than that of Bacon-Watts method, thus obtaining a more accurate knee-point. In addition, it is found that there exists linear relationship between the estimated knee-point and the battery life. Finally, the prediction results based on segmented model show that the prediction accuracy of segmented GA-SVR model is higher, and the MAPE is only 4.32%. The relevant results can provide some guidance for the reliability evaluation and product quality management of lithium-ion batteries, which is conducive to establishing more accurate models for battery life prediction.

Design of a pharmacokinetic/pharmacodynamic model for administration of low dose peripheral norepinephrine during general anaesthesia.

Intraoperative hypotension is a risk factor for kidney, heart and cognitive postoperative complications. Literature suggests that the use of low-dose peripheral norepinephrine (NOR) reduces organ dysfunction, yet its administration remains unstandardized. In this work we develop a pharmacokinetic (PK)/pharmacodynamic (PD) model of NOR and its effect on mean arterial pressure (MAP). From June 2018 to December 2021, we included patients scheduled for elective neurosurgery and requiring vasopressors for intraoperative hypotension management at Lariboisière Hospital, Paris. Low doses of NOR were administered peripherally, and successive arterial blood samples were collected to track its plasmatic concentration. We used a compartmental modelling approach for NOR PK. We developed and compared 2 models for NOR PD on MAP. Model comparison was done using Bayes information criteria. The resulting PK/PD model parameters were fitted over the entire population and linked to age, weight, height and sex. We included 29 patients (age 52 [46-64] years, 69% female). NOR median time to peak effect on MAP was 74 [53-94] s. After bolus administration, MAP increased by 24% (15-31%). A 2-comparment model with depot best captured NOR PK. NOR PD effect on MAP was well represented by both Emax and Windkessel models, with better results for the former. We found that age, height and weight as well as history of smoking and hypertension were correlated with model parameters. We have developed a PK/PD model to accurately track norepinephrine plasma concentration and its effect on MAP over time, which could serve for target-controlled infusion.

Minimal model identification of drum brake squeal via SINDy

AbstractThe industrial standard in the design and development process of NVH(Noise Vibration Harshness) characteristic of brakes is the application of Finite Element(FE) models with a high number of degrees of freedom in the range of one or several millions. Nevertheless, parallel experimental investigations are still indispensable. On the other hand, minimal models with, due to the inclusion of the self-excitation process, at least two degrees of freedom are well known to be capable to explain qualitatively phenomena as instability of the desired non-vibrating solution or limit cycle oscillation but are in general very inaccurate in predicting the dynamics of a specific real brake. This is because the underlying physical assumptions are already too restrictive and model parameters (especially those referring to nonlinearities) are widely unknown. To overcome this problem, the data-driven modeling approach SINDy(Sparse Identification of Nonlinear Dynamics) is applied to identify appropriate nonlinear functions for a brake squeal minimal model. A problem thereby is the limited database. It turns out that the naive implementation of the method yielding the lowest possible residuum does not necessarily provide physically meaningful models and results, respectively. Instead, a constrained model that incorporates physical knowledge is used to robustly identify parameters and reproduce realistic dynamic behavior. Thereby, several appropriate models with coexisting limit cycles and stationary equilibrium are identified. In particular, it was found that the angular position of the brake drum has a significant influence on the model parameters and therefore must be taken into account in a model with long-term validity.

Intervertebral disc creep behaviour through viscoelastic models: an in-vitro study

The intervertebral disc (IVD) is a complex biological structure that ensures the spine strength, stability, mobility, and flexibility. This is achieved due to its biphasic nature which is attained by its solid phase (annulus fibrosus) and fluid phases (nucleus pulposus). Hence, the IVD biomechanical response to long-term loads, which is critical as it affects hydration, and nutrients-water transport influencing disc height reduction, has been further explored and mathematically modelled in this paper. An in-vitro study was performed on seven human lumbar spine specimens (L4-5), to assess if the classical rheological models and Nutting's power law can model in a simple way the intermediate characteristics between solid and fluid of the IVD. Creep tests were conducted by applying a static compression load of 500 N for 15 min. A correlation analysis was done (Pearson, p < 0.05) between the model parameters and the maximum value of Disc Height Reduction, followed by a linear regression analysis. In summary, the long-term IVD mechanical behavior was modeled in a simple way, emphasizing that yet there is no mathematical certainty about this mechanical behavior. Hence, a future aim might be to develop intervertebral disc prostheses capable of replicating only the disc mechanical response, without necessarily considering the microscopic-level biological drivers. Therefore, a future goal is to fully understand and model the disc's mechanical response toward the design of new disc prostheses that would consider only the macroscopic aspect, without considering the biological aspects.

Impact of theoretical uncertainties on model parameter reconstruction from GW signals sourced by cosmological phase transitions

Different computational techniques for cosmological phase transition parameters can impact the gravitational wave (GW) spectra predicted in a given particle physics model. To scrutinize the importance of this effect, we perform large-scale parameter scans of the dynamical real-singlet extended Standard Model using three perturbative approximations for the effective potential; the MS¯ and on shell schemes at leading order, and three-dimensional thermal effective theory (3D EFT) at next-to-leading order. While predictions of GW amplitudes are typically unreliable in the absence of higher-order corrections, we show that the reconstructed model parameter spaces are robust up to a few percent in uncertainty. While 3D EFT is accurate from one-loop order, theoretical uncertainties of reconstructed model parameters, using four-dimensional standard techniques, remain dominant over the experimental ones even for signals merely strong enough to claim a detection by LISA. Published by the American Physical Society 2024

A fault diagnosis method of hot spots for photovoltaic clusters based on model parameters

AbstractUtilizing the direct current‐side electrical data resources of the photovoltaic power generation system on the FusionSolar platform, this study investigates the impact of hot spot faults on the output characteristics of photovoltaic strings and proposes a hot spot fault diagnosis method based on time series waveform characteristics. By analyzing the mechanisms of hot spot generation and evolution, as well as the characteristic differences in I–V curves and time series compared to other types of faults, the waveform variation patterns of hot spots in current and voltage time series are obtained. A function form suitable for hot spot fault waveform characteristics in time series graphs is constructed, and fault diagnosis feature vectors are extracted. Combining field operation and maintenance experience, a fuzzy reasoning fault diagnosis system is established to determine the causes and estimate the severity of hot spot faults. Experimental results indicate that hot spot faults have unique and corresponding variations in the current/voltage time series waveforms of the string output. The constructed function form can clearly represent the waveform variation patterns, and the established fuzzy reasoning system can achieve effective and reliable diagnosis of hot spot faults.

Dynamics Modeling and Parameter Identification for a Coupled-Drive Dual-Arm Nursing Robot

A dual-arm nursing robot can gently lift patients and transfer them between a bed and a wheelchair. With its lightweight design, high load-bearing capacity, and smooth surface, the coupled-drive joint is particularly well suited for these robots. However, the coupled nature of the joint disrupts the direct linear relationship between the input and output torques, posing challenges for dynamic modeling and practical applications. This study investigated the transmission mechanism of this joint and employed the Lagrangian method to construct a dynamic model of its internal dynamics. Building on this foundation, the Newton-Euler method was used to develop a dynamic model for the entire robotic arm. A continuously differentiable friction model was incorporated to reduce the vibrations caused by speed transitions to zero. An experimental method was designed to compensate for gravity, inertia, and modeling errors to identify the parameters of the friction model. This method establishes a mapping relationship between the friction force and motor current. In addition, a Fourier series-based excitation trajectory was developed to facilitate the identification of the dynamic model parameters of the robotic arm. Trajectory tracking experiments were conducted during the experimental validation phase, demonstrating the high accuracy of the dynamic model and the parameter identification method for the robotic arm. This study presents a dynamic modeling and parameter identification method for coupled-drive joint robotic arms, thereby establishing a foundation for motion control in humanoid nursing robots.

Facility Location Problem: Modeling Joint Disruptions Using Subordination

We study the facility location problem with disruptions where the objective is to choose a set of locations that minimizes the sum of expected servicing and setup costs. Disruptions can affect multiple locations simultaneously and are caused by multiple factors like geography, supply chain characteristics, politics, and ownership. Accounting for the various factors when modeling disruptions is challenging due to a large number of required parameters, the lack of calibration methodologies, the sparsity of disruption data, and the number of scenarios to be considered in the optimization. Because of these reasons, existing models neglect dependence or prespecify the dependence structures. Using partially subordinated Markov chains, we present a comprehensive approach that starts from disruption data, models dependencies, calibrates the disruption model, and optimizes location choices. We construct a metric and a calibration algorithm that learns from the data the strength of dependence, the number of necessary factors (subordinators), and the locations each subordinator affects. We prove that our calibration approach yields consistent estimates of the model parameters. Then, we introduce a variant of the standard approach to the underlying optimization problem, which leverages partially subordinated Markov chains to solve it quickly and precisely. Finally, we demonstrate the efficacy of our approach using twelve different disruption data sets. Our calibrated parameters are robust, and our optimization algorithm performs better than the simulation-based algorithm. The solutions from our model for disruptions have lower costs than those from other disruption models. Our approach allows for better modeling of disruptions from historical data and can be adapted to other problems in logistics, like the hub location, capacitated facility location, and so on., with joint disruptions. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2023.0103 .

Understanding Drug Exposure and Trichuris trichiura Cure Rates: A Pharmacometric Approach for Albendazole-Ivermectin Co-medication in Tanzania and Côte d'Ivoire.

Trichuriasis caused by the human whipworm Trichuris trichiura poses a significant public health concern. Albendazole-ivermectin co-medication is currently the most effective treatment. Studies conducted in Tanzania and Côte d'Ivoire unveiled differences in efficacy for albendazole-ivermectin combination therapy in both countries. A pharmacometrics approach was used to assess co-medication and study population effects on the pharmacokinetics of the two main metabolites of albendazole. An exploratory exposure-efficacy analysis was also carried out to investigate relationships between exposure measures and the egg reduction rate. Pharmacokinetic data from studies in Tanzania and Côte d'Ivoire in adolescents (aged 12-19 years) were included in the pharmacometric analysis. Participants received a single dose of either albendazole 400 mg alone or in combination with ivermectin 200 µg/kg. A pharmacometric analysis was performed to investigate the potential effects of the study population and co-administered ivermectin on the apparent clearance of the metabolites of albendazole. Non-linear mixed-effects modeling was conducted with MonolixSuite 2023R1. The pharmacokinetic exposure measures derived from simulations with individual model parameters were used in the exploratory-exposure response analysis. Pharmacokinetic profiles were best described by a two-compartment model for albendazole sulfoxide and a one-compartment model for albendazole sulfone, with a transit compartment and linear elimination. While no co-medication effect was found, apparent clearance of albendazole sulfoxide (albendazole sulfone) in the Tanzanian study population was 75% (46%) higher than that in the Côte d'Ivoire study population. Exposure-efficacy response analyses indicated that peak concentration and the time-above-exposure threshold were associated with the egg reduction rate. Study population but not co-administered ivermectin showed an effect on apparent clearance of albendazole sulfoxide and albendazole sulfone. Polymorphisms in drug-metabolizing enzymes and host-parasite interaction may explain this result. Difference in drug exposure did not explain the disparate efficacy responses in Tanzania and Côte d'Ivoire. Peak concentration and time-above-threshold were exposure measures associated with the egg reduction rate. Further studies evaluating genetic and resistance patterns in various regions in Africa are warranted.

BCDA: A blockchain-based dynamic auditing scheme for intelligent IoT

Intelligent Internet of Things (IoT) provides services for machine learning applications by collecting real-time data, where data correctness crucially impacts the training accuracy of models. Data owners host incremental real-time data in the cloud and use blockchain as a public platform for data query, usage, and certification. Malicious attackers may manipulate data through methods like data poisoning, compromising model parameters. Blockchain-based data integrity verification (auditing) schemes can effectively detect and monitor such activities. However, managing the storage demands for massive on-chain data certificates poses significant challenges. Existing compressed certificate storage strategies, such as the Merkle-Hash tree (MHT), incur significant on-chain storage costs because the auxiliary paths and the entire tree structure must be uploaded during verification, resulting in an O(n) storage overhead. Polynomial commitments offer a promising alternative by reducing storage costs to hundreds of bytes. However, current polynomial commitment methods lack effective support for incremental data updates, which are essential for real-time applications. In this paper, we propose a novel polynomial commitment-based dynamic integrity verification scheme that implements cumulative commitment updates for the first time. Our scheme includes a new data block insertion and deletion strategy that maintains the original order of data blocks within polynomial commitments. Theoretical security analysis confirms the robustness of the proposed scheme, while simulation results demonstrate that it meets the efficiency requirements for blockchain-based distributed data integrity verification mechanisms.

Study on bending fatigue performance of recycled aggregate backfill subgrade

Recycled aggregate (RA), as a backfill subgrade material, has strong reproducibility and environmental protection, which cannot only effectively reduce resource consumption and environmental pollution but also achieve recycling of resources. Therefore, a study on the bending fatigue performance of RA backfill subgrade is proposed. Based on linear elastic state, softening state, and damage accumulation state, the variation law of bending fatigue damage variables is analyzed, and the stiffness of RA under cyclic load is calculated. According to the correlation between fatigue damage corresponding to two different load links and the constitutive relationship of RA, the identification results of bending fatigue damage state based on RA backfilling subgrade is obtained. The advantages of the fatigue damage model of RA are analyzed, and the fatigue life equation is established based on the damage evolution equation. Strain, stiffness modulus, asphalt saturation, and asphalt mixture adjustment coefficient are selected as model parameters to establish the fatigue damage model of RA. The flexural bearing capacity of the double-reinforced rectangular section is calculated, and the flexural fatigue performance of RA backfill subgrade is analyzed. The test results show that the high stress level in this method leads to a sharp decline in the fatigue life of the specimen, and the influence of fatigue damage gradually appears, which is helpful to improve the durability and safety of the subgrade structure. The range of change shows a small range, which is close to the reduction coefficient result, indicating that this method has high reliability in analyzing the bending fatigue performance of RA backfill subgrade.